Product management is a delicate balance between vision, execution, and adaptability. Over the past year, I have had the chance to work as a fractional product manager working on chatbot solutions for education and teacher monitoring with Quest Alliance, and have had a chance to navigate this dynamic space, engaging with stakeholders, running experiments, and identifying key areas for growth. This post captures what worked, what could be improved, and the lessons learned from my journey.

What was done

• Visioning Workshop
• Field Visits
• Innovations like UDISE, Simplified Dashboards

What worked well

1. Co-Creation and Collaboration enabled by the Visioning Workshop
2. Field Visits
3. Innovations like UDISE, Simplified Dashboards

What Was Done Well

1. Co-Creation and Collaboration enabled by the Visioning Workshop – The two-day Chatbot Vision Planning Workshop brought together state leads, functional teams, and leadership to shape the product’s future. The approach emphasized:
   • On day 1 – Grounding participants in the current state of chatbot developments – This was done by showcasing working demos to set a foundation for further innovation and Encouraging exploration of external chatbot solutions to spark new ideas. This allowed for a realistic wish list – of innovative implementation inspired by others as well clear articulation of innovation in the space based on gaps seen across multiple products.
   • Structuring hands-on activities to ensure state-specific product roadmaps branching out of these common requests of innovation and innovative implementations.
   • By the end of the workshop, there was alignment on strategic pillars such as innovation, enablement, mainstreaming, and personalization—paving the way for a clear roadmap. While  getting sign-offs took time across different functions, it allowed the central leadership, M&E teams, and multiple state teams to come together and frame a common strategy.
2. Enhancing Teacher Engagement Through Chatbots – The teacher monitoring bot proved to be an effective tool for reminders and real-time feedback, offering a smoother experience compared to static Google Forms. Another blog will talk more about this.
3. Innovations and innovative implementations
   • UDISE validation – almost every student knows their UDISE code. We were able to reduce errors and standardise entry of States, Districts, Blocks, Clusters, School names using this approach.
   • LLM to clarify doubts – this made the engagement and doubt clarification more human for students
   • Simplified dashboards clearly calling out goals, and removing the need for team members at different levels to infer from a dashboard
   • Mailing dashboards – these on email to the team so that everyone had a pulse on the data
   • Research and Stats to see how many used shared phones, how many said they were on the bot and while they actually weren’t allowed us to get a better sense of possible engagement
   • Tracking engagement – by writing the user’s response to a nudge to a sheet allowed us to understand how we could track and engage users who engaged with our content.
   • Field visits and action research – to understand the actual on-ground impact of our bot and content

Opportunities for Improvement

1. Addressing Teacher Fatigue and Engagement Barriers

While the chatbot made data collection easier, teachers are known to have struggled with limited time and competing priorities to consistently participate.

Possible solutions:

• Embedding value-added nudges (e.g., providing immediate useful resources in return for feedback).
• Offering recognition or small incentives for regular participation. Implementing recognition or modest incentives for consistent participation can help sustain engagement among regular users and encourage a focus on the quality of their contributions, independent of others’ participation levels. Combined with the value added nudges, we should be able to counter the “Overjustification effect” where external rewards diminish intrinsic motivation over time.
• Ensuring less intrusive, more intuitive and byte-sized workflows for data submission.

2. Strengthening Localised Inclusivity

Feedback suggested that students and teachers not only preferred content in their native language (e.g., Gujarati), but also with local context. While this is a known challenge in the space, the challenge was to generate content at scale which allowed for this, especially, if the ask for practical, hands-on exercises were in high demand.

Potential enhancements – we are planning to test a feature called the Open Challenge Framework where students who have consumed a particular content can generate challenges leveraging their local context and share it with us. We then review it for factual correctness, preserving the local flavour, validate that the local context is neither too niche, nor lacking enough context for others outside to solve the probelm and share it as weekly challgnes. This allows application of knowledge and measuring

3. Closing the Feedback Loop for Teachers

One recurring issue was the lack of follow-up on teacher responses. While data collection was robust, teachers needed to see tangible outcomes from their input.

Actionable steps:

• Implementing personalized feedback reports for teachers.
• Highlighting key insights from teacher responses in structured updates.
• Creating community forums for educators to discuss and share best practices.

4 Automated Data cleaning

One key learning was that I had to run multiple runs of data cleaning. This was time consuming and high on effort. Over time, I have seen opportunities to automate some of this, and it seems to be a valuable learning.

---

Key Takeaways for Product Managers

1. Balance Vision with Execution: While long-term strategy is crucial, iterative execution and real-world testing ensure sustained impact. Leverage
2. Engage Stakeholders Early and Often: The chatbot’s success was driven by collaborative input from multiple teams, ensuring that solutions were tailored to real needs.
3. Keep the User at the Center: Understanding teachers’ and students’ motivations helped refine product decisions for greater engagement.
4. Leverage Data for Continuous Learning: Every experiment—successful or not—offered critical insights that shaped future iterations.
5. Iterate, Iterate, Iterate: The journey of product management never ends. Every challenge is an opportunity to refine and improve.

---

Conclusion

The experience of developing and iterating chatbot solutions for education has been both insightful and challenging. While significant strides were made in engagement, feedback collection, and innovation, there remain opportunities to enhance teacher motivation, content accessibility, and the overall value proposition of chatbots in education. By continuously learning, adapting, and refining, product managers can drive meaningful, scalable impact in EdTech and beyond.